Both social scientists and policy makers have been concerned about determinants and consequences of adolescent sexuality. In response to that concern, several recent projects have collected extensive data on adolescent sexuality and related variables. This proposal requests support to analyze existing, high quality adolescent sexuality data using unique methodologies and a newly developed conceptual framework. Data from 334 sibling pairs are contained in the ADSEX data collected at the University of North Carolina in five waves. In addition to the sibling pairs, friendship pairs and classmate pairs can also be constructed from these data. These links give a broad range of environmental sources to search for possible explanations of the antecedents of adolescent sexuality. The family environment, the social environment, and the school environment can be studied using equivalent and complementary analyses. The specific goals include searching each of these environments for the antecedents to adolescent sexual behavior. The longitudinal nature of the dataset can be used to separate selection from peer influences; and the within family nature of the dataset can be used to separate common environmental from unique environmental influences. The long-range goals include the development of a more complete understanding of the environmental influences on adolescent sexual behavior for use by both social scientists and policymakers. Such understanding will help identify the appropriate arenas to which adolescent sexuality and adolescent pregnancy programs should be directed.